<h2>Must know</h2>
<ul>
<li><strong>AI-generated code is causing real problems in production</strong>: Turns out, AI isn’t quite the coding genius we were promised. ChatGPT is spitting out code with deprecated APIs and non-existent property wrappers, leading to widespread build failures and state management nightmares in SwiftUI apps. Meanwhile, AI-generated SwiftUI UI, while initially promising, is structurally unsound, causing velocity degradation and requiring full refactors that take longer than building from scratch, according to Rebeloper. It seems the AI can write code fast, but understanding context, architecture, and maintainability is still a human superpower. <a href="https://app.daily.dev/feed-by-ids?id=nqrakn9b2&amp;id=yIClekHIh&amp;id=NDJH1JzvV&amp;id=STWE5tRXe&amp;id=UAphs9F5B" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>OpenAI is getting desperate for cash, pushing ads and hardware</strong>: OpenAI is rolling out ads in ChatGPT’s free tier, aiming for billions in revenue to fund its AGI ambitions, a move CEO Sam Altman once called a ‘last resort’. This comes as their enterprise LLM market share has plummeted from 50% to 27%, with Anthropic now leading at 40%, according to TechCrunch. To further diversify, OpenAI is reportedly developing AI-powered earbuds, codenamed ‘Sweet Pea,’ designed by Jony Ive’s team, to compete with Apple’s rumored AI pin. It seems the dream of AGI is expensive, and they’re willing to turn ChatGPT into a billboard to pay for it. <a href="https://app.daily.dev/feed-by-ids?id=DQh9IAXNl&amp;id=ep7p1hnQy&amp;id=9fjLFvZxt&amp;id=ArnlILqap&amp;id=vOBU57mgd&amp;id=fgglWe8Qn" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>Mandiant weaponizes NTLMv1 to force its retirement</strong>: Mandiant, a Google subsidiary, just dropped rainbow tables that can crack Net-NTLMv1 authentication in under 12 hours on a cheap consumer PC. This isn’t just for kicks; it’s a deliberate move to force organizations to finally ditch the ancient, insecure NTLMv1 protocol, which is still lurking in many legacy Windows environments. Microsoft has already removed support in Windows Server 2025, but Mandiant’s move makes exploitation trivial, putting the onus on IT teams to disable it immediately or face easy credential theft. <a href="https://app.daily.dev/posts/2q2phjvWS" target="_blank" rel="noopener nofollow">Read more</a></li>
</ul>
<h2>Good to know</h2>
<ul>
<li><strong>AI agents are getting smarter, but still need human babysitters</strong>: New benchmarks like AssetOpsBench show even top AI agents like GPT-4.1 only achieve 72.4% execution scores in industrial asset management, falling short of deployment readiness, while researchers are using LLMs to generate 40+ exploits for zero-day vulnerabilities in hours, not weeks. <a href="https://app.daily.dev/feed-by-ids?id=C3HmVtQKc&amp;id=WkT9zuzY1" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>TypeScript is eating the world, Rust is for the brave, and Python is still the AI king</strong>: TypeScript adoption surged 65% year-over-year, dominating modern app development across front-end, back-end, and AI products, while Python remains essential for AI and data workflows, and Rust gains traction for security-critical systems, according to Tech With Tim. <a href="https://app.daily.dev/posts/RJIgj9I8E" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>Microsoft’s January update is a dumpster fire for Windows users</strong>: Microsoft’s January 2026 security update is causing Windows apps to freeze when saving to cloud storage, breaking remote desktop authentication, and preventing proper shutdowns, impacting both consumers and enterprises. <a href="https://app.daily.dev/feed-by-ids?id=KR0gCuX2r&amp;id=1mNBVIaM1&amp;id=YPaWZCLLg" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>AI is transforming software development, but human skills are more critical than ever</strong>: AI coding tools are shifting the developer role from coding to orchestrating AI agents, requiring strong communication, specification, and architectural understanding, as AI excels at implementation but struggles with context and complex problem-solving. <a href="https://app.daily.dev/feed-by-ids?id=sifgyKGyd&amp;id=aKVj0i6Bm&amp;id=xiDxzfJJR&amp;id=ruSCTD0Fh&amp;id=S37WCeeOE" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>The EU is cracking down on big tech, pushing for digital sovereignty</strong>: The EU is entering a ‘heavy enforcement’ phase for its Digital Markets Act and Digital Services Act, aiming to reduce dependence on US tech giants and reshape digital infrastructure with a focus on open-source and open standards, according to TNW Weekly Briefing. <a href="https://app.daily.dev/feed-by-ids?id=SMoUF4qcl&amp;id=nH5xYsaji&amp;id=AirINdKGE&amp;id=Kn5AMUlFL&amp;id=TV1Su1iZZ" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>New Linux kernel features are making life better for Intel and AMD users</strong>: Linux 7.0 is bringing limited non-root access to Intel-Speed-Select, improved GPU temperature monitoring for Intel graphics, and firmware updates for Intel discrete GPUs on ARM64 and RISC-V, while a new kernel patch significantly cuts wakeup latency on modern Intel Xeon CPUs. <a href="https://app.daily.dev/feed-by-ids?id=uDY1GxTuI&amp;id=39acsU5oQ&amp;id=LPWAW20mV&amp;id=NWx6tDyeo" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>AI is making JavaScript optimization a breeze</strong>: LLMs are proving highly effective at optimizing JavaScript performance, reducing turnaround time from hours to minutes and transforming game framerates from 20-30 FPS to near 60 FPS, even handling complex bit-shifting algorithms, according to Joshua Morony. <a href="https://app.daily.dev/posts/NttZ4o4xB" target="_blank" rel="noopener nofollow">Read more</a></li>
</ul>
<h2>Bullshit police</h2>
<ul>
<li><strong>AI is not replacing developers, it’s just making them review more code</strong>: Despite the constant hype, AI isn’t replacing developers; it’s just generating a ton more code that humans then have to review. Claude Code, for instance, generates 6.4 times more code than humans, increasing pull request volume by 98% and review time by 91%, according to a hands-on experiment. So much for AI making our lives easier, now we’re just drowning in its ‘defensive’ code. <a href="https://app.daily.dev/feed-by-ids?id=BrFQI7FHb&amp;id=BrFQI7FHb" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>The ‘AI-native’ transformation is mostly just marketing fluff</strong>: Everyone’s talking about ‘AI-native’ platforms and ‘vibecoding’ as the future, but it sounds like a lot of hot air. Companies like Emergent are raising hundreds of millions for ‘vibe-coding’ platforms that promise to let non-programmers build full-stack apps, while others like Strapi are launching ‘AI-native’ CMSs. Meanwhile, the actual impact on productivity and ROI for enterprise AI projects remains elusive, with many failing to move beyond pilot phases due to infrastructure complexity, according to DDN’s survey. <a href="https://app.daily.dev/feed-by-ids?id=AllrxKqHW&amp;id=AllrxKqHW&amp;id=McILqauKu&amp;id=f5IhxeEO5&amp;id=ing2AgJbz&amp;id=k12AAd1Ej&amp;id=e3oUC0g8q" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>AI is making us dumber, one prompt at a time</strong>: Apparently, relying too much on AI for basic tasks like writing emails is going to make our critical thinking skills ‘atrophy,’ according to some unnamed experts. And let’s not forget ChatGPT’s penchant for repeating myths as truth due to ‘training noise.’ So, while AI might be great for generating code, it seems it’s also excellent at making us less capable and more susceptible to misinformation. Maybe we should all go back to pen and paper before our brains turn into mush. <a href="https://app.daily.dev/feed-by-ids?id=JB9XOabfe&amp;id=oVxKIKmTT&amp;id=eU5DB6Li1&amp;id=eU5DB6Li1" target="_blank" rel="noopener nofollow">Read more</a></li>
</ul>
